Successfully reported this slideshow.

English as an indo european language text


Published on

Published in: Technology, Education
  • Be the first to comment

  • Be the first to like this

English as an indo european language text

  1. 1. 12 English as an Indo-European Language Philip BaldiLike most of the more than 5,000 languages in the world, English belongs to a lan-guage family, that is, a group of languages that are related to each other geneticallyand share a common ancestry. The “genes” they share are inherited linguistic featureswhich have been transmitted through time over the history of the languages in ques-tion. The notion of a language family is founded on the observation that two or morelanguages may contain features of lexicon (vocabulary), phonology (sound), morphol-ogy (word structure), and syntax (grammar) which are too numerous, too fundamental,and too systematic to be due to chance, to general features of language design (typol-ogy), or to borrowing through contact. The language family to which English belongsis known as the Indo-European (IE) language family, and the common ancestor fromwhich the Indo-European languages derive is called Proto-Indo-European (PIE). Thesubgroup within Indo-European to which English belongs is Germanic, specificallyWest Germanic. As we begin our exploration of English as an IE language, we will first spend sometime discussing the methods by which languages are classified genetically, how thesemethods help us to separate linguistic structures that are inherited from those whichare not, and how they are used to access the past, including the preliterary past, oflanguages such as English. How do we know that languages share “genetic material,” and are therefore to begrouped within the same language family? We begin with a few simple illustrationswith languages which will be familiar to most readers. Everyone knows that the “Romance” languages (such as French, Italian, andSpanish) are all in some way descended from Latin. What this means is that theRomance languages are all “sister” languages, and that they stem from a commonancestor, thereby forming a genetic group (more specifically a subgroup). We knowthis on independent factual grounds, based on the documented history of the RomanEmpire and its spread throughout early Europe. But even in the absence of historicalrecords tracing the spread of the Romans and their language in its various forms, weA Companion to the History of the English Language Edited by Haruko Momma and Michael Matto© 2008 Blackwell Publishing Ltd. ISBN: 978-1-405-12992-3
  2. 2. 128 Philip BaldiTable 12.1 One through ten in some Romance languages Italian French Spanish Portuguese Rumanian Latinone uno un uno um unu unustwo due deux dos dois doi duothree tre trois tres tres trei tresfour quattro quatre cuatro quatro patru quattuorfive cinque cinq cinco cinco tint quinquesix sei six seis seis sase sexseven sette sept siete sete sapte septemeight otto huit ocho oito optu octonine nove neuf nueve nove nao novemten dieci dix diez dez dzate decemTable 12.2 One through ten in some Germanic languagesEnglish Dutch German Swedish Yiddishone een eins en eynstwo twee zwei två tsveythree drie drei tre drayfour vier vier fyra firfive vijf fünf fem finfsix zes sechs sex zeksseven zeven sieben sju zibneight acht acht åtta akhtnine negen neun nio naynten tien zehn tio tsenwould arrive at the same conclusion of linguistic relatedness through the comparisonof the modern languages. Consider, for example, the lower numerals in selected“major” Romance languages (table 12.1), written in standard orthography (which mayobscure features of pronunciation). Of course the existence of similarities among these five Romance languages is easyto explain. They share a common ancestor language (Latin), and have inherited thelower numerals directly from this source; i.e., the words are “cognates” and the lan-guages are “sisters.” But there are equally compelling data from languages whoseancestor can only be inferred because, unlike Latin, it was never written down. Con-sider the modern members of the Germanic subgroup (table 12.2). Despite the obvious relatedness and common ancestry in the Romance and Ger-manic examples just cited, such connections are not always obvious. And even whenit is convincingly established that the languages in question are in some sort of his-torical relationship, it is by no means an easy step to determine what the ancestor
  3. 3. English as an Indo-European Language 129might have looked like, when and where it was spoken, or what other languages mightbe related, perhaps more distantly (i.e., as “cousins” rather than “sisters”). The Latin-Romance connection is deceptively simple because of what we know about the linesand stages of transmission between the historical end points (Latin and Italian, forexample). The Germanic case is somewhat more difficult because of the absence of anattested ancestral language (there are older Germanic languages, such as Gothic orOld Icelandic, but these are not proto-systems). Nonetheless, the evidence for related-ness among these languages is just as powerful as with the Romance languages. Wejust don’t have a written ancestor. Are such resemblances enough to prove a genetic relationship among languages?Are we forced to conclude from these displays of vocabulary in a limited field (here,lower numerals) that the languages in each group are derivable from some commonancestor? Surely there are other explanations available to account for the likenesses –borrowing through language contact, for example. Languages exchange vocabularywithout regard for family membership; need and prestige are the two primary factorswhich govern the borrowing process. The languages which make up the respectiveRomance and Germanic subgroups have been in close cultural and geographicalcontact for millennia, so might it not be conceivable that they all just borrowed thenumbers 1–10 from one or the other of them, or perhaps some other language? For the lexicon to be used even as a preliminary guide to possible genetic relation-ships, we need more examples of potential cognates than a few (admittedly impressive)sets of numerals. In particular we need vocabulary items which, like the numerals,are part of the “core” vocabulary, i.e., words which are unlikely to have been borrowed,and which exist in sufficient quantity to exclude the possibility of chance (see table12.3). Like the numerals, these words come from deep in the core of the lexicon. Theyare not technical terms, like computer or fax, nor do they represent culturally transport-able items such as pizza or sushi. And there are countless numbers of sets like them,eliminating the factor of chance. The only reasonable way to account for these simi-larities is to treat the words as cognates, and to assume that they are derived from acommon source. We call that source language “Proto-Germanic.”Table 12.3 Some “core” Germanic vocabularyEnglish Dutch German Swedish Yiddishlove liefde Liebe ljuv “sweet” libeto live leven leben leva lebnto fly vliegen fliegen flyga flienhand hand Hand hand hanthouse huis Haus hus hoyzmy, mine mijn mein min maynmother moeder Mutter mo(de)r mutername naam Name namn nomen
  4. 4. 130 Philip Baldi Classifying languages based on vocabulary similarities represents only the first stepin the historical process. To complete the task, we have to take a closer look at theproperties of the words we have assembled to determine the degree of systematicitywhich holds across the languages. If the languages are indeed related (as we knowthese to be), the correspondences in vocabulary should be matched by systematic cor-respondences in phonology and morphology as well (syntax is somewhat more prob-lematic). The principle of regularity is the cornerstone of the comparative method,by which linguists reconstruct the parent language and its intermediate stages basedon the comparative analysis of the descendant languages. So, if say English andSwedish are related, and if there is a correspondence such that Eng. /m/ correspondsto Swed. /m/ in a given phonetic environment, then it should be the case for every /m/(see the examples for “my,” “mother,” “name”); likewise for /l/ (see “love,” “live,” “fly”)or for /v/ and /b/ (see “love” and “live” in English and German). As we work out thedetails of such correspondence sets we make inferences about the ancestral sound,which in the first two cases would be postulated as *m and *l (with the * designatinga hypothetical reconstructed segment). For every set of words in which Swed. /m/corresponds to Eng. /m/ in a given phonetic environment, we claim that both derivefrom a common proto-sound *m in Proto-Germanic. The same principle holds as wework to progressively more distantly related languages, such as Latin and (Old)English, or Greek and Sanskrit, using the oldest available data as we work backwardsin time, all the way to PIE. Needless to say the correspondences become less and lessobvious with deeper time spans and the need for auxiliary explanatory mechanismssuch as analogy and secondary sound change increases, but the method is sophisticatedenough that it can reveal correspondences over millennia of distance in first attesta-tion, say between Old English (ca. 600 ce) and Ancient Greek (ca. 800 bce) or Hittite(ca. 1750 bce). The Indo-European Language FamilyThe term “Indo-European” refers to a family of languages which by about 1000 bcewere spoken over a large part of Europe and parts of southwestern and southern Asia(see figure 12.1). The dating and location of a unified PIE is controversial in many respects, but themost widely held opinion among specialists puts the protolanguage in the area of thePontic-Caspian steppes north of the Black and Caspian Seas at about 3500 bce, afterwhich it began to diversify into the descendant subgroups through phases and stageswhich are matters of debate (more than a few locales and time horizons have beenproposed). Though the concept of “Indo-European” is linguistic, the term is originallygeographic, referring to the location of the easternmost (India) and westernmost(Europe) languages at the time the family was securely identified in the eighteenthand nineteenth centuries. In comparison with some of the other 250–300 languagefamilies of the world, the IE family is relatively small. It contains about 140 languages
  5. 5. Publishers Note: Permission to reproduce this image online was not granted by the copyright holder. Readers are kindly requested to refer to the printed version of this chapter.Figure 12.1 Distribution of Indo-European Languages, circa 500–1000 bce (from P. Baldi (2002) The Foundations of Latin. New York: Walter de Gruyter;map 1, p. 37). Reprinted with permission
  6. 6. 132 Philip Baldi(many extinct), more than 90 of which belong to Indo-Iranian; these 140 or so lan-guages are classified into 11 subgroups, one of which is Germanic, where English islocated. By contrast, the Austronesian language family of the Pacific has some 800languages in a large number of subgroups, and the Bantu family (Africa) has as manyas 400 languages. Of course it is important to distinguish the number of languagesin a family from the number of speakers, or the geopolitical importance of the lan-guages in question (as evidenced by their status as second languages, or as a linguafranca). By these latter criteria the Indo-European family, specifically through thecolonial and global languages such as French, Spanish, and especially English, has aunique standing among the language families of the world. The family tree represents graphically some of the more important and recognizablemembers of the IE family (figure 12.2). We offer here a few words about each sub-group, its dating, and its overall importance for our understanding of PIE and itshistory. AnatolianCompletely extinct, the Anatolian languages were unknown until archeological exca-vations in Bogazköy, Turkey in the early twentieth century uncovered the royal ˇarchives of the ancient Hittite city of Hattusas. The original trove of about 10,000 ˇ ˇclay tablets (now about 25,000), dating from the seventeenth to the thirteenthcenturies bce, was deciphered from its cuneiform script and shown to be representingan Indo-European language now called Hittite. The discovery, classification, andeventual detailed analysis of the Anatolian languages, but especially Hittite, hasimpacted IE studies significantly. Before Hittite, PIE was reconstructed with a “look”that resembled the older IE languages, in particular Baltic, Slavic, Greek, Latin,and Sanskrit. But Hittite, though demonstrably older, does not share a numberof structural features with the “classical” IE languages, and in many cases displayscharacteristics which can be shown to significantly predate those in other IElanguages. Two of the more famous of these archaisms were the existence of severalsounds (called “laryngeals”) that had been lost in the other subgroups, often leavinga trace; and the absence of the “classical” three-way gender system (masculine-feminine-neuter) in substantives in favor of a two-way animate-inanimate system.Accounts of differences such as these between Hittite and the other IE languageshave challenged the traditional look of reconstructed PIE and its chronology, prompt-ing some scholars to view the Anatolian languages as sisters, rather than daughters,of PIE, with both descending from a more remote protolanguage called“Indo-Hittite.” Indo-IranianThis subgroup contains two closely related subdivisions, namely Indic (Indo-Aryan)and Iranian.
  7. 7. Faroese Icelandic Norwegian Danish Swedish Polish Ukrainian Czech Slovak Belarusian Russian West North Germanic East North Germanic Gothic West Slavic East Slavic North Germanic East Germanic Lithuanian SLAVIC GERMANIC Old Prussian Latvian South Slavic BALTIC West Germanic Slovenian Old Church Slavic Serbo-Croatian Bulgarian Old English Old Frisian Old Low Franconian Old Saxon Old High German Middle English Frisian Middle Dutch Middle Low German Middle High German TOCHARIAN English Dutch Afrikaans Low German German Yiddish PROTO-INDO-EUROPEAN Flemish Tocharian A Tocharian B CELTIC INDO-IRANIAN Armenian Insular Continental Iranian Goidelic Brittanic Celtiberian Gaulish Albanian West Iranian East Iranian (Old Persian) (Old Avestan)Irish Scots Gaelic Welsh Breton Illyrian Thracian Manx Cornish ITALIC Kurdish Farsi Pashto Yahgnobi Phrygian Parthian Khotanese Indic Latin-Faliscan Oscan-Umbrian HELLENIC ANATOLIAN Sanskrit Vedic Latin Faliscan Pakrits North/West South/East Romance Lydian Hittite Hindi-Urdu Pali Aeolic Doric Arcado-Cyprian Mycenaean Luwian Punjabi Portuguese Romanian Attic-Ionic French Spanish Italian GreekFigure 12.2 Major Indo-European branches and languages
  8. 8. 134 Philip BaldiIndic (Indo-Aryan)The languages of the Indic group are classified into three historical periods, namelyOld Indic (1500–600 bce), Middle Indic (600 bce–1000 ce), and Modern Indic(since 1000 ce). The most ancient language is Vedic, an archaic form of Sanskritwhose oldest documents are dated by some to about 1200–1000 bce, though othersconsider them to be older. A closely related form of Vedic is Classical Sanskrit, which -.was codified in the work of the grammarian Panini ca. 500 bce, and in which several -important literary texts are written. The oldest Middle Indic texts are in Pali (sixthto fifth century bce), followed by the Asoka inscriptions (ca. 250 bce) and some ´Jainist religious writings from about the same period. Modern Indic is one of thelargest and most heterogeneous of the IE subgroups, with perhaps as many as ninetydifferent languages. Among the best known of them are Hindi-Urdu, Marathi,Punjabi, and Gujurati.IranianAncient Iranian has two important representatives. The chief one of these is OldAvestan (also known as Gathic Avestan), dating from about 600 bce, possibly earlier.The second important member of Ancient Iranian is Old Persian, a Western Iranianlanguage, which may date to as early as 500 bce. Western Middle Iranian is repre-sented by Middle Persian (Pahlavi) and Parthian, while the Eastern Middle Iranianlanguages are Sogdian, Khotanese, Khorasmian, and Bactrian. Modern descendants ofIranian are Modern Persian (Farsi), Pashto, and Kurdish. GreekAlso known as Hellenic, the Greek branch contains some of the oldest testaments ofIndo-European. Attested inscriptionally from as early as the eighth century bce,Greek has textual monuments in the Homeric epics the Iliad and the Odyssey, whichmay be as old as 800 bce. Even older than these are the Linear B tablets from Crete,Pylos, and other ancient locales which represent a form of Greek called Mycenaeanand may be from as far back as the fourteenth century bce. The two principal subdi-visions are between South/East Greek (comprising Attic-Ionic, Arcado-Cyprian, andMycenaean), and North/West Greek (comprising Aeolic and Doric). The main dialectof Greek is Attic, the literary language of Athens in which standard Classical Greekliterature was composed. Standard Modern Greek developed from Attic-Ionic. ItalicThe Italic subgroup of Indo-European consists of many genetically connected lan-guages from ancient Italy which share certain distinctive characteristics. Thereare two main Italic subdivisions. The more important of the two, Latin-Faliscan, is
  9. 9. English as an Indo-European Language 135represented chiefly by Latin, one of the most important IE languages and arguablythe most important language in the development of Western Civilization. Latin isidentifiable in some short inscriptions from the seventh century bce, though the firstcontinuous literature stems from the third century bce. Faliscan is known only frominscriptions, the oldest of which dates to the early seventh century bce. Latin survivesin the modern Romance languages, which developed from spoken varieties of thelanguage in various parts and at different times and social circumstances in the historyof the Roman Empire and beyond. The best known of the Romance languages areItalian, French, Spanish, Portuguese, Catalan, and Rumanian; less well known areDalmatian, Rhaeto-Romansch, Ladino, Sicilian, Sardinian, Occitan, and many otherlocal and social varieties. The second Italic subdivision is called Osco-Umbrian (alsoSabellic or Sabellian). There are no modern descendants of this branch, which com-prises Oscan (attested as early as the fifth century bce), Umbrian (perhaps as early as300 bce), South Picene (fifth to sixth century bce), and a number of fragmentarylanguages. Some classification schemes put Italic in a special subunity with Celticknown as “Italo-Celtic.” GermanicThe Germanic subgroup, which includes English among its members, is widespreadgeographically and is internally heterogeneous. The oldest attestations of Germanicare the Scandinavian Runic inscriptions, the oldest of which dates from the firstcentury ce. The Germanic languages are conventionally separated into three geo-graphic subdivisions. The first, East Germanic, contains only a single well-attestedlanguage, Gothic. Gothic is the language with the oldest continuous documents inGermanic, the biblical translation by Bishop Wulfila from around the second half ofthe fourth century ce. The second subdivision of Germanic is North Germanic, whoseprincipal representative is Old Icelandic (also called Old Norse). Apart from the Runicinscriptions, the oldest material in North Germanic comprises Norwegian and Ice-landic sagas and legal texts from the ninth century. Modern North Germanic lan-guages are Icelandic, Faroese, and Norwegian in one group, and Danish and Swedishin another. The final group, West Germanic, is the most expansive and internallydiverse of the Germanic languages; its descendants include German, Yiddish, Dutch,Flemish, Afrikaans, and English, with its many varieties worldwide. (See furtherenglish as a germanic language.) CelticThe languages of the Celtic subgroup are traditionally divided into two main geo-graphical sections, the Continental and the Insular. The Continental group, made upof Celtiberian (Hispano-Celtic), Lepontic, and Gaulish, is extinct. The oldest materialfrom the Continental group is from the sixth century bce. The Insular Celtic lan-guages show up materially somewhat later. Split into two groups, Goidelic and
  10. 10. 136 Philip BaldiBrittanic (Brythonic), the Insular languages are first attested in some Ogham Irishsepulchral inscriptions from around 300 ce. The Goidelic group is made up of Irish,Scots Gaelic, and the extinct Manx. Brittanic comprises Welsh, the most robust ofthe modern Celtic languages, Breton, spoken in Brittany (France), and the extinctCornish. Some classification schemes put Celtic in a special subunity with Italicknown as “Italo-Celtic.” TocharianDiscovered in archeological excavations around the turn of the twentieth century inChinese Turkestan, the two varieties of Tocharian (usually called simply “A” and “B”)have added modestly to the Indo-European base. The documents of the languages,mostly religious and some commercial, are relatively late, stemming from the periodof about 500–700 ce. BalticSometimes grouped with the Slavic languages to form a composite intermediatebranch called “Balto-Slavic,” the Baltic subgroup survives in two modern langu-ages, Lithuanian and Latvian (Lettish), which together make up the East Baltic sub-division. Many other Baltic languages have become extinct, including a languagecalled Old Prussian, which was spoken until the early eighteenth century and repre-sents the West Baltic subdivision. The oldest Baltic material, the Old PrussianBasel Epigram, dates to as early as 1369 ce, while the oldest Lithuanian texts stemfrom the early sixteenth century, and the oldest Latvian material is probably datableto 1585. SlavicOften grouped with Baltic as “Balto-Slavic,” the Slavic languages fall into three geo-graphical subdivisions. The first, South Slavic, comprises Bulgarian, Macedonian,Serbian, Croatian, Slovenian, and the extinct Old Church Slavic, in which the bulkof the oldest (tenth century) Slavic materials are written. The second Slavic subdivi-sion is West Slavic, which comprises Czech, Slovak, Polish, Kashubian, and someothers. And finally there is East Slavic, made up of Russian, Ukrainian, andBelarussian. ArmenianArmenian is first attested in religious documents and translations from the fifthcentury ce. It shows a great deal of influence from neighboring languages, includingGreek, Arabic, Syriac, and Persian, so much so in fact that it was first misclassifiedas a dialect of Iranian.
  11. 11. English as an Indo-European Language 137 AlbanianUnknown linguistically until the fifteenth century ce, Albanian shows a great dealof influence from neighboring languages such as Greek, Slavic, and Turkish, as wellas from Latin. This made its secure identification as a branch of Indo-European some-what problematic when the IE languages were being classified in the eighteenth andnineteenth centuries. The first document in Albanian is a baptismal formula from thefifteenth century. There are two principal dialects, Gheg and Tosk. Fragmentary languagesIn addition to the 11 major subgroups, there are also many apparently unaffiliatedlanguages which survive only in fragments such as glosses and sporadic inscriptions.These languages provide enough information to be classified as IE, but not muchbeyond that. Included among the fragmentary IE languages are Ligurian (northernItaly, possibly related to Celtic), Messapic (southern Italy, possibly connected withIllyrian), Sicel and Sicanian (Sicily), Venetic (northeastern Italy), Thracian (in the areaof modern Bulgaria and southern Romania), Phrygian (in the area of modern centralTurkey), Illyrian, from the Dalmatian coast area of the Adriatic), and several others. Aspects of the structure of PIEThe extensive comparison of the daughter languages and their analysis according tothe comparative method and other established methodologies has led to a protolan-guage that has been reconstructed in considerable detail. In this section we willidentify some of the more prominent features of reconstructed “classical” PIE, espe-cially those relevant for the history of English, largely omitting revisions, includinglaryngeals, based on Anatolian evidence. PhonologyTable 12.4 shows the correspondences between selected consonant and vowel seg-ments in several ancient IE languages and the oldest Germanic languages. Recon-structed PIE initiates the correspondences. Table 12.5 provides a few illustrative lexical reconstructions. (See furtherphonology: segmental histories.) MorphologyNominal and pronominal morphology“Classical PIE,” that is, the PIE reconstructed before the integration of Anatolianevidence into the protolanguage, is considered to be an inflectional (fusional) language
  12. 12. 138 Philip BaldiTable 12.4 Phonological correspondence among some PIE languagesPIE Hitt. Skt. Lat. Gk. Goth. OIc. OHG OEp p p p p f f f ft t t t t þ þ d þk k ´ s k k h(j) h h hkw ku k/c qu p/t/k hw/w hv hw/w hwb p b b b p p p/pf pd t d d d t t z tg k j g g k k k kgw ku g/j gu/u b/d/g qu kv q cw/kbh p bh f(b) ph b b b bdh t dh f(d) th d d t/d dgh k h h kh g g g ggwh ku gh/h f ph/th/kh w w w ws s s s s s s s sm m m m m m m m mn n n n n n n n nl l l/r l l l l l lr r r/l r r r r r rw/u w v v Ø w v w wy/i y y j h/z j Ø j g(y)a a a a a a a a æe e, a a e e i e e ei i i i i i i i io a/a ¯ a o o a a a æu u u u u u u u ua¯ a ¯ a ¯ a ¯ a/e ¯ ¯ o ¯ o ¯ o ¯ o ¯e¯ e ¯ a ¯ e ¯ e ¯ e ¯ a ¯ a ¯ – æo¯ a ¯ a ¯ o ¯ o ¯ o ¯ o ¯ o ¯ o ¯(Hitt. = Hittite; Skt. = Sanskrit; Lat. = Latin; Gk. = Greek; Goth. = Gothic; OIc. = Old Icelandic; OHG= Old High German; OE = Old English)in which case markers on nominals (nouns, adjectives, pronouns) indicate their gram-matical relationship to other words in a sentence, and mark gender and numberagreement among words in phrases. The protolanguage is traditionally reconstructedwith eight (occasionally nine) cases which indicate grammatical and semantic distinc-tions such as subjecthood, objecthood, direction towards, dislocation from, temporal-ity, exchange, possession, agency, and instrumentation. The cases are known as the
  13. 13. English as an Indo-European Language 139Table 12.5 Some PIE reconstructions, based on “core” vocabulary from IE languages Hitt. Skt. Lat. Gk. Goth. OIc. OHG OE PIEthree teri- tráyah . tre s ¯ treîs þreis þrır ¯ drı ¯ þrı ¯ *trei-seven sipta- ˇ saptá septem heptá sibun siau sibun seofon *septm ºcow wa-wa-(i)-× gáuh bos ¯ boûs ⊕ ky ´r chuo – cu w *g ou- .I – uk ahám ego ´ ego(n) ¯ ik ek ih ic *eg-foot pata-× ´ pat ¯ pedisº podósº fotus ¯ fotr ¯ fuoz fot ¯ *ped-heart kard- ⊕ cordisº kardía ¯ haírto ¯ hjarta herza heorte *kerd-sheep hawi-× ávih ovis o(w)is ⊕ – œr ouwi e owu ¯ *owi- . ˘× The form is Hieroglyphic Luwian.º The genitive case reveals the stem.⊕ The cognate form is not found in this language.Table 12.6 A sample noun declension (Lat. servus “servant”)Case Singular PluralNominative servus servı ¯Genitive servı ¯ servorum ¯Dative servo ¯ servıs ¯Accusative servum servos ¯Ablative servo ¯ servıs ¯Vocative serve servı ¯nominative, genitive, dative, accusative, ablative, locative, instrumental, and vocative.Number refers to the quantification of entities in a phrase; the protolanguage hadthree numbers (singular, dual, and plural), as well as three genders (gender is anunfortunate term which simply means a kind of noun class), namely masculine, femi-nine, and neuter. Adjectives followed the same pattern of inflection as nouns, andagreed in gender and number with their head noun. Pronouns are marked by theirown more-or-less unique endings. Latin provides a useful analog to the PIE system,though without the locative and instrumental (table 12.6). In the Latin sentence Marcus servum vocat “Marcus calls the servant,” Marcus’ roleas subject is marked by the ending -us and the servant’s role as object is indicated by-um. The order of the words is grammatically irrelevant (Latin, like PIE, usually putsthe verb at the end). When words occur as members of a constituent (word group),their membership is indicated by shared endings marking case, number, and gender, ¯¯ ¯ ¯ -as in velo cı equo “to the swift horse” [dative-singular masculine], malo rum animalium
  14. 14. 140 Philip Baldi“of the bad animals” [genitive-plural-neuter], or ab aliıs feminıs “from the other ¯ ¯ ¯women” [ablative-plural-feminine]. (See further history of englishmorphology.)Verbal morphologyPIE verbs are synthetically complex amalgamations of meaningful elements whichindicate grammatically and semantically significant categories. The PIE verb encodedtwo voices, active and mediopassive (voice reflects the role of the subject); a numberof tenses (tense locates the verbal action temporally: at least the present, imperfect,aorist, perfect, and possibly a future are usually postulated); and mood, which indi-cates the factual content of the utterance from the speaker’s point of view: at least theindicative, imperative, and optative moods are reconstructed, and occasionally thesubjunctive. Voice, tense, and mood markers are attached to stems indicating aspec-tual categories (e.g., whether the action is continuous or punctual), and the entirecomplex is indexed to the subject by means of person/number markers. Verbs can betransitive (i.e., they can govern an object as in “Mary sees Bill”) or intransitive (“Sarahwalks to school”), though there is no specific formal marking on the verb to distin-guish the transitive and intransitive types. Once again Latin can be instructive, though it is not a perfect replica of PIE: a – -verb form like am-a ba-t in rex ama bat “the king used to love” contains a stem form ¯ –(am-), which indicates the lexical meaning “love”; a mood marker (-a -), markingindicative (factual) mood; a tense/aspect marker (-ba-) which marks continuous pastaction; and finally a person/number/voice marker (-t), which indicates third personsingular in the active voice. If we change the example to the passive rex ama ba tur ¯ - -(am-a–-ba -t-ur) “the king used to be loved,” the marker of passivity is the final -ur; – - –in the plural reges ama bantur (am-a -ba-nt-ur) “the kings used to be loved” note that ¯¯the person/number marker is now -nt-. SyntaxFusional languages like PIE and many of its descendants (including Old English,though not to the same extent as Latin, Greek, or Sanskrit) have fundamentally dif-ferent syntactic patterns from languages like Modern English or French. The reasonhas much to do with word order, and the fact that a good deal of the syntax of fusionallanguages is conveyed in morphological expressions, such as case endings. In ModernEnglish, for example, the order of elements in a sentence is grammatically fixed: exceptin stylistically marked utterances such as “Bagels, I like,” the subject precedes theverb, and the object follows the verb in simple sentences. It is not grammatical to say“Him John sees” or “Sees him John” to mean “John sees him.” But in fusional lan-guages like PIE, word order is a stylistic, not a grammatical device. Latin is illustrativeagain: Marcus me vocat “Marcus calls me” represents the preferred (unmarked) order of ¯elements, but me vocat Marcus or Marcus vocat me have the same semantic value as ¯ ¯
  15. 15. English as an Indo-European Language 141Marcus me vocat. That is because the grammatical indication of subject (Marcus) and ¯object (me) is being carried by the endings, not the position of the words relative to ¯each other; furthermore, the verb vocat is indexed by the final -t to the third personnominal subject Marcus, and couldn’t possibly go with me. PIE (like Old English) was ¯dominantly verb-final (John him sees). Verb-final languages have certain properties suchas: they use postpositions (the world over); adjectives typically precede the noun theymodify (the proud winners), also true for genitives (Susie’s exam); comparative construc-tions have the order standard-marker-adjective (Louis than taller [= taller than Louis]);and relative clauses precede the noun they qualify (who teach English professors [= profes-sors who teach English]). (See further history of english syntax.) The ways in which many of these features of PIE descended into Germanic and onto English are discussed in english as a germanic language, in this volume. AcknowledgmentThanks to Richard Page and Aaron Rubin, who commented on an earlier version ofthis chapter. References and Further ReadingFortson, B. W., IV (2004). Indo-European Language Ramat, A. G. & Ramat, P. (eds.) (1998). The Indo- and Culture: An Introduction. Oxford: Blackwell. European Languages. London: Routledge.Mallory, J. P. & Adams, D. Q. (1997). Encyclopedia of Szemerényi, O. (1996). Introduction to Indo-European Indo-European Culture. London: Fitzroy Dearborn. Linguistics. 4th edn. Oxford: Oxford UniversityOnions, C. T. (1969). The Oxford Dictionary of Press. English Etymology. Oxford: Clarendon Press. Watkins, C. (2000). The American Heritage Diction-Pokorny, J. (1951–9). Indogermanisches etymologisches ary of Indo-European Roots. 2nd edn. Boston: Wörterbuch. Bern: Francke. Houghton Mifflin.